knitr::opts_chunk$set(echo = FALSE)
library(tidyverse)
library(DT)

setwd("C:/github/Dionex-HPLC-Data-Parser")
source("get_data_ftn.R")
source("calibrate_ftns.R")
source("apply_cal_ftns.R")
source("apply_dilution_factor_ftn.R")
source("remove_tbl_dups_ftn.R")

Lohse Lab Dionex Data Parser


Calibration Standard Analyte Concentrations

setwd("C:/github/Dionex-HPLC-Data-Parser")
path_to_file <- "data/DionexTest_raw.xlsx"

cal_vals <- suppressMessages(get_data(file_name = path_to_file)$cal_values)
run_data <- suppressMessages(get_data(file_name = path_to_file)$run_data)

datatable(cal_vals)


Dionex Run Output 

datatable(run_data)


Generate calibration curves

By default, a low and high calibration curve is created using standards 1-4 and 4-8

Notes: Removed standard 8 due to inaccurate analysis

### Generate table of calibration curves
curves <- data.frame(analyte =      c(rep("Chloride", 2), rep("Nitrate", 2), rep("Sulfate", 2)),
                     level =        rep(c("High", "Low"),3),
                     ex_std_from =  rep(c(1,5),3),
                     ex_std_to =    rep(c(3,8),3),
                     rmv_std =      rep(c("Std 8"), 6)) 

curves_all <- curves %>% split(1:nrow(curves)) %>% 
                         map(~ generate_cal_curve(
                                  analyte=.x$analyte,                    
                                  cal_values = cal_vals, 
                                  run_df = run_data,
                                  remove_std = .x$rmv_std,
                                  exclude_std = paste0("Std ", seq(from=.x$ex_std_from, to=.x$ex_std_to)))) %>%
                         bind_rows() %>% 
                         cbind(curves, .)

# Apply calibration to determine analyte concentrations
run_conc_tbl <- apply_cal_to_run(cal_curves = curves_all, run_output = run_data)

# Remove irrelevant columns
run_conc_tbl2 <- run_conc_tbl %>% select(No., SampleID, Analyte, Area, Concentration, RNote) %>% mutate_if(is.numeric, ~round(., 3))


Check Standard Accuracy

check_df <- run_conc_tbl2 %>% filter(grepl("CHECK_", SampleID))

unique_ions <- unique(run_conc_tbl2$Analyte)

check_tbl_df <- check_df %>% select(No., SampleID, Analyte, Concentration)

check_tbl_df$Accuracy <- round(check_tbl_df$Concentration/as.numeric(gsub("CHECK_", "", check_tbl_df$SampleID,"_")),3)*100

check_tbl_df$Note <- ifelse(check_tbl_df$Accuracy < 95 | check_tbl_df$Accuracy > 105, "WARNING", "")
  
datatable(check_tbl_df, rownames = F,
            options = list(
              pageLength = 100))


Data Tables

datatable(run_conc_tbl2, rownames = F,
          extensions = 'Buttons', options = list(
            pageLength = 100,
            dom = 'Bfrtip',
            buttons = 
              list('copy', 'print', list(
                extend = 'collection',
                buttons = c('csv', 'excel', 'pdf'),
                text = 'Download'
            ))))
# Apply dilution multipliers
run_conc_tbl3 <- run_conc_tbl2 %>% split(1:nrow(run_conc_tbl2)) %>% map(apply_dil_factor) %>% bind_rows()  

datatable(run_conc_tbl3, rownames = F,
          extensions = 'Buttons', options = list(
            pageLength = 100,
            dom = 'Bfrtip',
            buttons = 
              list('copy', 'print', list(
                extend = 'collection',
                buttons = c('csv', 'excel', 'pdf'),
                text = 'Download'
            ))))



Finalized Data Table

#Isolate the unknown sample data

ions <- unique(run_conc_tbl3$Analyte)

final_data <- NA

for(i in 1:length(ions)) {
  unknowns_tbl <- run_conc_tbl3 %>% filter(Analyte == ions[i]) %>%
                              filter(SampleID != "blank") %>%
                              filter(SampleID != "Blank") %>% 
                              filter(!grepl("CHECK_", SampleID)) %>%
                              filter(!grepl("Std", SampleID)) %>%
                              filter(!grepl("std", SampleID))
  
  # Add dilution factor column
  unknowns_tbl$Dilution_Factor <- strsplit(unknowns_tbl$RNote, "factor x") %>%
          do.call(rbind, .) %>%
          as.data.frame() %>%
          select(V2)
  
  unknowns_tbl$Dilution_Factor <- suppressWarnings(as.numeric(unlist(unknowns_tbl$Dilution_Factor)))
  
  output <- unique(unknowns_tbl$SampleID) %>% map(~remove_dups(uniqID = ., 
                                                               tbl = unknowns_tbl)) %>% bind_rows()
  
  if(is.na(final_data)) {
    final_data <- output
  } else {
    final_data <- left_join(final_data, output) #It is expected to get console message for this join
  }
}

#Rearrange dataframe
col_DilF_names <- colnames(final_data)[grepl("DilF",colnames(final_data))]
col_Note_names <- colnames(final_data)[grepl("Note",colnames(final_data))]
  
final_data <- final_data[c("Run_Pos", "SampleID", ions, col_DilF_names, col_Note_names)]

# Show data table in RmD
datatable(final_data, rownames = F,
          extensions = 'Buttons', options = list(
            pageLength = 100,
            dom = 'Bfrtip',
            buttons = 
              list('copy', 'print', list(
                extend = 'collection',
                buttons = c('csv', 'excel', 'pdf'),
                text = 'Download'
            ))))
# Save datatable as .csv and .rds
setwd("C:/github/Dionex-HPLC-Data-Parser")
write.csv(final_data, "Data/R_parsed_run_data.csv")
saveRDS(final_data, "Data/R_parsed_run_data.rds")